Clinical Significance of Mitral Regurgitation After Acute Myocardial Infarction
Background Mitral regurgitation (MR) may complicate acute myocardial infarction (MI). However, it is not known whether mild MR is an independent predictor of post-MI outcome.
Methods and Results The study cohort consisted of 727 Survival and Ventricular Enlargement Study patients who underwent cardiac catheterization, including left ventriculography, up to 16 days after MI. Left ventriculograms were analyzed for diastolic and systolic volumes, global left ventricular sphericity, extent of wall motion abnormality, and endocardial curvature. The presence of MR was related to the risk of developing a cardiovascular event during 3.5 years of follow-up. MR was present in 141 patients (19.4%). Severe (3+) MR was present in only 2 patients. Patients with MR were more likely to have a persistently occluded infarct artery (MR versus no MR, 27.3% versus 15.2%; P=.001). Although the ejection fractions were similar, MR patients had larger end-systolic and end-diastolic volumes and more spherical ventricles than patients without MR. Sphericity change from diastole to systole was also significantly reduced in MR patients. Patients with MR were more likely to experience cardiovascular mortality (29% versus 12%; P<.001), severe heart failure (24% versus 16%; P=.0153), and the combined end point of cardiovascular mortality, severe heart failure, or recurrent myocardial infarction (47% versus 29%; P<.001). The presence of MR was an independent predictor of cardiovascular mortality (relative risk, 2.00; 95% CI, 1.28 to 3.04).
Conclusions Mild MR is an independent predictor of post-MI mortality. As such, it adds important information for risk stratification of post-MI patients.
Mitral regurgitation is known to be a frequent complication of AMI. When present, it may exhibit a broad range of severity, from clinically evident and hemodynamically obvious to clinically silent and detected only as an incidental finding on catheterization or Doppler echocardiography. Indeed, when it is sought by Doppler,1 MR has been reported to occur in up to 39% of patients with MI. Papillary muscle dysfunction and associated dysfunction of the underlying ventricular wall are thought to be the most common cause of MR in post-AMI patients,2 3 and MR has generally been identified as a more frequent complication of inferior than anterior infarction. Although severe MR may place an additional hemodynamic stress on the LV, its prognostic significance independent of LV function has been controversial. Barzilai et al4 found that AMI patients with a murmur suggestive of MR had a 12-month mortality of 36% compared with 15% for patients without an MR murmur. However, correction for differences in baseline variables indicated that the presence of an MR murmur was not an independent predictor of outcome. In contrast, Lehmann et al5 found that MR present on left ventriculography within 7 hours of MI was an independent predictor of survival at 1 year. Tcheng et al6 also found that moderately severe to severe MR appeared to be a likely independent predictor of impaired survival.
The purpose of the present study was to analyze ventriculographic data from post-MI patients with LV dysfunction to determine the LV wall motion, volume, and geometric correlates of MR and whether the presence of MR is an independent predictor of clinical outcome.
The SAVE trial was a double-blind, placebo-controlled, randomized multicenter trial that demonstrated that chronic therapy with the ACE inhibitor captopril enhances survival and clinical outcome in survivors of AMI with LV dysfunction. Patients were 21 to 80 years old and had a radionuclide LVEF ≤40%. Patients were excluded from the study if they had overt congestive heart failure at the time of randomization. All patients underwent a prerandomization physical examination during which the presence or absence of a murmur of MR was specifically elicited on the case report form. Randomization to placebo or captopril occurred 3 to 16 days (average, 4 days) after a documented AMI. After randomization to placebo or captopril (titrated to a maximum dose of 50 mg TID), patients were followed for 2 to 5 years (mean, 3.5 years). All surviving patients underwent a repeat radionuclide LVEF measurement during the last months of follow-up. Complete aspects of protocol design as well as the final results of the SAVE study have been published.7 8
Although SAVE did not restrict or mandate specific aspects of clinical management before the qualifying MI and randomization, in patients with overt ischemia, cardiac catheterization was required before randomization. This requirement, in addition to clinical preferences across centers, led to a clinically directed cardiac catheterization being obtained between the qualifying MI and randomization in 1301 of the 2231 patients. Moreover, if catheterization identified a need for revascularization therapy, the study required that revascularization be completed before randomization. Participating SAVE centers submitted 990 angiograms from these 1301 patients for inclusion in the core laboratory analyses. The present study is based on the group of patients whose cardiac catheterizations included left ventriculography of sufficient quality to analyze the presence of MR. Thus, 743 patients (75% of 990) underwent left ventriculography based on clinical indications and local practice in each clinical site, and the other 147 patients underwent catheterization without left ventriculography being performed. However, poor angiographic quality led to 16 studies being excluded because MR could not be assessed. Thus, MR could be graded in 727 patients (98% of 743).
Analysis of Left Ventriculograms
Analysis of LV wall motion, presence of MR, and coronary anatomy was carried out by one of three experienced invasive cardiologists (G.A.L., G.C.F., or S.C.S.). The method of ventricular analysis has been described previously.9 The presence and extent of mitral regurgitation were graded on a standard scale (0 to 4).10 The left ventriculograms at end diastole and at end systole were traced onto a transparent film and digitized at a resolution of 10 points per millimeter by use of a digitizing tablet interfaced to a personal computer. When extrasystoles were present, care was taken to analyze a cardiac cycle at least two beats after the last extrasystole. Computer-assisted analysis of the traced left ventriculograms provided the following: (1) LV volumes calculated by the area-length method11 in the 359 patients (49.4% of 727) in whom calibration for magnification was possible; (2) determination of LV shape by use of the sphericity index,12 an index of overall LV shape based on calculating ventricular volume and dividing it by a hypothetical spherical volume generated by use of the longest axis of the LV as the diameter of a sphere. Calculation of the sphericity index does not require correction for magnification; (3) analysis of wall motion at 100 chords by the centerline method.13 For each left ventriculogram, the percentages of the diastolic LV circumference that were dyskinetic, akinetic, hypokinetic, normal, and hyperkinetic were calculated on the basis of comparisons with normal values obtained on 76 patients without coronary disease or a history of MI; and (4) analysis of regional LV shape by measurement of the endocardial curvature. Each ventricular silhouette was converted to polar coordinates and expressed as a Fourier series by use of a fast Fourier transform.14 Curvature (K), the reciprocal of radius (R) of curvature (K=1/R), was calculated at each of the 100 centerline wall-motion chords. Curvature was corrected for ventricular size by multiplying each value by the endocardial circumference of the LV silhouette being analyzed. This resulted in a dimensionless descriptor of local geometry. Positive curvature denotes outward convexity and negative curvature denotes outward concavity of the endocardial silhouette.
Analysis of Coronary Angiograms
Coronary arteries were analyzed on the basis of standard, multiple-view angiograms. Multiple segments (proximal, mid, and distal) in each coronary artery and, when present, bypass grafts were analyzed by the angiographic analysis scheme previously described.9 Each individual coronary artery segment was measured by digital electronic caliper and analyzed for percent reduction in luminal diameter and antegrade and collateral flow. These findings were used to define infarct artery patency in patients whose infarct could be electrocardiographically localized and whose infarct artery could be identified. Occlusion of the infarct artery required the presence of grade 0 or 1 antegrade flow and grade 0 or 1 collateral flow as defined by the TIMI investigators.15 The patency status of the infarct artery was further modified by the results of revascularization performed between the SAVE MI and randomization, as previously described.9 Patients who underwent percutaneous transluminal coronary angioplasty of the infarct artery before randomization had their patency status redefined on the basis of the final results of the angioplasty. Patients with occluded infarct arteries who underwent successful surgical bypass of an occluded infarct artery before randomization were also considered to have patent infarct arteries.
The analysis of baseline characteristics in mutually exclusive categories was examined by χ2 statistics, and continuous variables were compared by one-way ANOVA and expressed as mean±SD. Multivariate analyses to determine whether MR of any severity was an independent predictor of cardiovascular mortality were carried out by the proportional-hazards model reported by Cox.16 Survival tables were prepared by the Kaplan-Meier method.17 These multivariate analyses were carried out on the overall group of patients who underwent left ventriculography and necessarily excluded quantitative ventriculographic variables that were not available on all patients.
Baseline Clinical Characteristics
Compared with the 1504 SAVE study patients who were not part of this database, the 727 patients who underwent cardiac catheterization with left ventriculography and formed a part of this database were generally younger, were more likely to have received thrombolytic therapy or mechanical revascularization therapy, had a slightly higher LVEF, and were less likely to have been in Killip class >1 early after MI (Tables 1⇓ and 2⇓).
Within the population clinically selected for cardiac catheterization, a murmur of MR detected by physical examination was present in 5% of patients within 24 hours of SAVE enrollment. In contrast, MR by LV angiogram was present in 141 (19.4%) of these 727 patients; 9% of patients with angiographic MR had clinically audible MR reported. The angiographic severity of MR was 1+ in 106 (14.6%) and 2+ in 33 (4.5%). Severe MR (3+) was present in only 2 patients. Patients with MR and those without MR had similar LV filling pressures (LV end-diastolic pressure, 23±9 versus 22±9 mm Hg, P=.110). Patients with MR were older and more likely to have a history of multiple infarctions (Table 2⇑). There was a trend toward more women in the MR group. They also were less likely to have received thrombolytics during their AMI and less likely to be receiving a β-adrenergic receptor blocking agent before randomization into SAVE. Patients with MR were more likely to have sustained an inferior infarction, a combined anterior-inferior infarction, or an AMI of indeterminate location than patients without MR. LVEF demonstrated no significant differences between patients with and without MR.
Coronary angiography was available for all 727 patients. Patients with MR had more severe coronary disease than did patients without MR (MR versus no MR: single-vessel disease, 36.9% versus 52.2%; multivessel disease, 63.1% versus 47.8%; P<.001). In addition, patients with MR were also more likely to have a persistently occluded infarct artery at the time of enrollment into SAVE (MR versus no MR, 27.3% versus 15.2%; P=.001).
Left ventriculography demonstrated that patients with MR had greater end-systolic and end-diastolic volumes than did patients without MR (Table 3⇓). In addition, the LVs of patients with MR were significantly more spherical in both systole and diastole than were the LVs of patients without MR. Sphericity change from diastole to systole was significantly reduced in MR patients. Quantitative wall motion analyses demonstrated that there were no significant differences in the percentage of the diastolic LV circumference that was dyskinetic or akinetic. However, patients with MR demonstrated a significantly larger hypokinetic segment than did patients without MR.
In diastole, the anterobasal and anterior regional endocardial curvatures were not significantly different in patients with and without MR. In contrast, apical diastolic curvature was less in patients with MR than in those without MR. In addition, the inferior wall diastolic endocardial curve was significantly less convex outward in patients with MR than in those without MR.
Analysis of systolic endocardial curvature demonstrated significant differences between patients with and without MR in all regions except for the LV apex. Patients with MR demonstrated a less concave anterobasal segment during systole than did patients without MR. The anterior segment was less convex and the inferior segment also was less concave in patients with MR than in those without MR.
Patients with MR demonstrated a significantly worse overall prognosis than did patients without MR for all the clinical end points tested (MR versus no MR: cardiovascular mortality, 29% versus 12%, P<.001; severe heart failure, 24% versus 16%, P=.0153; and the combined end point of cardiovascular mortality, severe heart failure, or recurrent myocardial infarction, 47% versus 29%, P<.0001) (Table 4⇓, Figure⇓).
Independent Predictors of Cardiovascular Mortality
A multivariate model was constructed that included as independent variables those characteristics previously reported in the SAVE population to be correlated with clinical outcome, as well as other important clinical descriptors of the study population. These included age, sex, history of hypertension, diabetes, prior infarction, thrombolytic therapy at the time of the SAVE infarction, LVEF, number of diseased vessels, infarct artery patency, presence of MR, and treatment group assignment (captopril or placebo). Hypertension, LVEF, number of diseased vessels, and the presence of MR were found to be independent predictors of cardiovascular mortality (Table 5⇓).
Effect of Captopril Therapy
In the overall ventriculography group of 727 patients, there was no significant difference in cardiovascular mortality between patients treated with captopril or placebo (14.8% versus 16.1%, P=.635). Likewise, there were no significant differences in cardiovascular mortality between captopril- or placebo-treated patients in the subgroups of patients with MR (captopril versus placebo, 26.8% versus 31.4%; P=.542) or without MR (captopril versus placebo, 12.0% versus 12.3%; P=.905).
Mitral regurgitation often develops during the course of AMI. For example, Barzilai et al4 found an MR murmur present on admission in 9% of patients with acute MI and present sometime during the hospitalization in 20%. Heikkila et al18 described a murmur suggestive of MR in 55% of MI patients, and Bhatnagar and Yusuf19 used color-flow Doppler to confirm that 9% of post-MI patients who had a murmur compatible with MR did, indeed, have MR. More recently, angiographic studies reported by Lehmann et al5 and Tcheng et al6 reported incidences of post-MI MR of 13% and 17.9% of patients within hours of infarction, respectively. The present study differs from the above-mentioned angiographic studies principally because of the timing of diagnostic catheterization an average of 4.2 days after the index infarction and selection for LV dysfunction. However, the overall incidence of MR in the present study, 19.4%, is similar to that reported in the angiographic studies above.
Patients with MR were, on average, older than patients without MR, and there was a trend toward an excess of women in the MR group. Similar data on the demographics of MR after AMI have been presented by Lehmann,5 Tcheng,6 and Barzilai.4 The SAVE study excluded very few patients with severe MR. Therefore, it is interesting to note that even mild MR was more frequently associated with Killip class >1 during the first 72 hours of MI. This finding may reflect the early clinical effects of the abnormalities of LV function and geometry discussed below.
LV Function and Geometry
Mittal et al2 speculated that papillary muscle dysfunction alone is insufficient to cause MR after MI and that an underlying wall motion abnormality is needed. Others have supported this hypothesis3 by reporting that in dogs, hypokinesia of the ventricular segment overlying the papillary muscle, which leads to retraction of the mitral leaflets toward the apex, is a sufficient condition to produce MR. In the present study, patients with MR demonstrated the same total degree of wall motion abnormality as patients without MR. Thus, the presence of MR and the poorer prognosis of MR patients is not due to a larger infarct size in patients with MR, and more subtle differences in LV structure and function must be considered.
In contrast to earlier studies, the present study measured the dimensionless variables global LV sphericity and endocardial curvature in a large cohort of patients. On average, patients with MR had larger and more spherical ventricles than did patients without MR. In addition, the shape change from diastole to systole, which normally results in a more ellipsoidal systolic ventricular shape, was reduced in patients with MR. This association of mild, or “functional,” post-AMI MR with increased LV sphericity was reported as a clinical observation by Perloff and Roberts20 in 1972. In 1973, Vokonas et al21 analyzed 25 patients with chronic, hemodynamically significant MR of various causes, calculated an index of ventricular geometric eccentricity, and found that patients with MR had a more spherical chamber geometry and that the shape change from diastole to systole was diminished in patients with compensated as well as decompensated MR. Gould et al22 also analyzed patients with a variety of cardiac lesions and calculated wall stress in patients with MR. These investigators reported that an increase in LV sphericity will lead to an alteration in myocardial fiber orientation and an increase in meridional wall stress. Kono et al23 suggested that LV sphericity is an essential component of functional MR in patients with congestive heart failure and cardiomyopathy. In this admittedly different cohort, Kono et al reported that increased LV sphericity was the only difference between patients with MR and those without MR. Indeed, the importance of LV geometry in preserving normal mitral valvular function had previously been emphasized by Perloff and Roberts20 and, more recently, by Sabbah and coworkers.24 The latter reported that chamber enlargement alone is insufficient to account for functional MR and that a distortion of LV shape leading to an alteration in the angle at which chordal structures are tethered to the mitral leaflets is an important component. The present study strongly supports these hypotheses.
However, the present study also measured LV endocardial curvature independent of global shape and volume. Patients with MR had significant regional systolic endocardial curvature abnormalities compared with those without this abnormality. In particular, patients with MR had a less concave systolic anterobasal wall. There were abnormalities of inferior wall curvature as well. The inferior wall normally has a slightly outward concave endocardial curvature at end systole. Patients with MR and inferior infarctions may have systolic bulging of the inferior wall, leading to the less outwardly concave curvature reported. Thus, the global LV shape abnormalities in patients with MR are independent of LV volumes and are the result of the multiple regional endocardial curvature abnormalities described above. Nevertheless, it is difficult to interpret regional curvature abnormalities in a mixed population of anterior and inferior MI patients. In previous studies, we have shown that anterior MIs generally have increased outward concavity in the anterobasal region early after MI, and this area flattens out with remodeling. Moreover, curvature values for the inferior wall may be confusing, because anterior and inferior MIs have directly opposite effects. An anterior MI will tend to have a markedly accentuated outward concavity in this region, whereas inferior MIs will tend to have a markedly diminished concavity or even outward convexity in this region. Thus, the summary data are more properly interpreted by noting that there are marked abnormalities of global LV shape that are related to regional differences in curvature remote from the apex, which is markedly abnormal in all patients independent of the presence of MR.
Earlier studies have emphasized that MR in ischemic heart disease is most prominent in patients with inferior or inferoposterior infarctions.25 This has been postulated to be due to the particular vulnerability of the posterior papillary muscle to its blood supply. In the present study, a preponderance of inferior or inferoposterior infarcts among patients with MR was observed.
The present analyses demonstrate that patients with MR had more severe coronary disease, with a greater incidence of three-vessel disease and a lower incidence of single-vessel disease, than did patients without MR. These findings differ from those of Lehmann et al,5 who found no significant differences in extent of coronary disease when comparing patients with or without MR. In contrast, Tcheng et al6 reported that 18% of patients without MR and 33% of patients with MR had three-vessel coronary artery disease. Both of these series differ from the present study in that catheterizations were all performed within hours of the acute infarction.
Acute pharmacological or mechanical reperfusion therapy has been reported to reverse acute severe MR in most,26 27 28 although not all,29 reports. The present study shows that patients with MR were less likely to have received thrombolytic therapy during the acute MI and less likely to present with a patent infarct artery.
One of the most important findings of this study is that MR, when angiographically mild, is often clinically unsuspected. Indeed, only 9% of patients with angiographically evident MR had an audible murmur reported by experienced observers. Nevertheless, although MR was generally mild and often clinically unsuspected, its presence in the post-MI patient was associated with a highly significant increase in clinical risk. These results are consistent with those of earlier investigators.4 5 6 30 Published studies demonstrate that the presence of MR, whether by physical examination, Doppler ultrasound, or left ventriculography, portends a poor post-MI prognosis. However, although it is reasonable to suppose that severe MR may carry with it a poor prognosis due to the severe hemodynamic load it imposes on the post-AMI ventricle, the present study emphasizes the importance of even trace MR.
Earlier studies of MR often did not grade its severity. Maisel et al30 defined MR by physical examination and found it to be present in 17% of their patients with MI. Patients with a systolic murmur had a 1-year mortality of 18% compared with 10% for those without a regurgitant murmur. Barzilai et al4 also reported a 1-year mortality of 36% for those patients with an MR murmur detected on admission compared with 15% for those without a murmur. However, MR was not graded in an objective fashion.
In contrast, Lehmann et al5 performed LV angiography and graded MR in 206 patients. MR was mild in almost all patients. Even patients with mild MR demonstrated a nearly fourfold increase in mortality at 1 year, and multivariate analyses suggested that MR was an independent predictor of survival. Tcheng et al6 reported that patients with no MR, mild MR, and severe MR had a stepwise increase in overall mortality. However, mild MR did not appear to be an independent predictor of survival.
The importance of mild MR could be addressed in the SAVE study, because SAVE entry criteria excluded patients with severe MR. Furthermore, extensive additional data were available regarding demographics, clinical care during hospitalization, physical findings, coronary anatomy, and most importantly, LV volumes and geometry. Univariate analyses demonstrated that patients with MR had larger LV volumes, more spherical LV chambers, and more severe coronary disease than patients without MR. A multivariate analysis that included coronary anatomy, LVEF, and extent of coronary disease found MR to be an independent predictor of cardiovascular mortality.
There are several potential reasons why angiographically mild MR is an independent predictor of cardiovascular outcome. The present study convincingly demonstrates that patients with larger, more geometrically abnormal LVs are more likely to have MR, despite similar systolic function as defined by LVEF. We postulate that mild MR is a marker of these geometric abnormalities. Thus, although mild MR may not present a severe hemodynamic load due to simple valvular regurgitation, we hypothesize that the noninfarcted myocardium of these ventricles is laboring under a severe hemodynamic load because of the marked geometric abnormalities present in MR patients. There also may be additional abnormalities of LV function or geometry that were undetected by the present single-plane analyses, for which MR therefore is a marker.
The assumption that mild MR detected on the catheterization table represents an inconsequential hemodynamic load may be flawed. Qualitative angiographic grading may be a far from ideal method of assessing the severity of MR. In subjects with 1+ to 2+ MR, the regurgitant volume may exhibit a broad range.31 Furthermore, the present study cannot exclude increasing severity of MR over time, particularly in ventricles subject to LV remodeling and dilatation. Finally, the effect of exercise and ischemia on the severity of MR also may affect its hemodynamic severity, particularly over time.
Effect of Captopril
In the present subgroup of SAVE patients who underwent left ventriculography, a significant benefit of captopril therapy on cardiovascular mortality was not observed. This finding is different from the overall results of the SAVE study and may be due to selection criteria for cardiac catheterization and for performance of left ventriculography in the catheterization subgroup.
Angiographically mild MR after infarction is often clinically unrecognized. However, its presence on the left ventriculogram of a post-MI patient with LV dysfunction is a marker of a larger, more geometrically distorted LV and is associated with a significant increase in the risk of subsequent cardiac death. Thus, careful analysis of the left ventriculogram will permit more accurate risk stratification of the post-MI patient.
Selected Abbreviations and Acronyms
|AMI||=||acute myocardial infarction|
|LV||=||left ventricular, left ventricle|
|SAVE||=||Survival and Ventricular Enlargement Study|
The SAVE study was funded by a grant from the Bristol-Myers-Squibb Institute for Pharmaceutical Research.
Reprint requests to Gervasio A. Lamas, MD, Division of Cardiology, Mount Sinai Medical Center, 4300 Alton Rd, Miami Beach, FL 33140.
- Received October 15, 1996.
- Revision received January 31, 1997.
- Accepted February 5, 1997.
- Copyright © 1997 by American Heart Association
Barzilai B, Gessler C, Pérez JE, Schaab C, Jaffe AS. Significance of Doppler-detected mitral regurgitation in acute myocardial infarction. Am J Cardiol. 1988;61:220-223.
Mittal AK, Langston M, Cohn KE, Selzer A, Kerth W. Combined papillary muscle and left ventricular wall dysfunction as a cause of mitral regurgitation: an experimental study. Circulation. 1971;44:174-180.
Kono T, Sabbah HN, Rosman H, Alam M, Jafri S, Stein PD, Goldstein S. Mechanism of functional mitral regurgitation during acute myocardial ischemia. J Am Coll Cardiol. 1992;19:1101-1105.
Barzilai B, Davis VG, Stone PH, Jaffe AS. Prognostic significance of mitral regurgitation in acute myocardial infarction. Am J Cardiol. 1990;65:1169-1175.
Lehmann KG, Francis CK, Dodge HT, and the TIMI Study Group. Mitral regurgitation in early myocardial infarction. Ann Intern Med. 1992;117:10-17.
Tcheng JE, Jackman JD, Nelson CL, Gardner LH, Smith LR, Rankin S, Califf RM, Stack RS. Outcome of patients sustaining acute ischemic mitral regurgitation during myocardial infarction. Ann Intern Med. 1992;117:18-24.
Moyé LA, Pfeffer MA, Braunwald E. Rationale, design and baseline characteristics of the Survival and Ventricular Enlargement Trial. Am J Cardiol. 1991;68:70D-79D.
Pfeffer MA, Braunwald E, Moyé L, Basta L, Brown EJ, Cuddy TE, Davis BR, Geltman EM, Goldman S, Flaker GC, Klein M, Lamas GA, Packer M, Rouleau J, Rouleau JL, Rutherford J, Wertheimer JH, Hawkins CM, on behalf of the SAVE Investigators. Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction. N Engl J Med. 1992;327:669-677.
Lamas GA, Flaker GC, Mitchell G, Smith SC, Gersh BJ, Wun CC, Moyé L, Rouleau JL, Rutherford JD, Pfeffer MA, Braunwald E, for the Survival and Ventricular Enlargement Investigators. Effect of infarct artery patency on prognosis after acute myocardial infarction. Circulation. 1995;92:1101-1109.
Grossman W. Profiles in valvular heart disease. In: Grossman W, Baim DS, eds. Cardiac Catheterization, Angiography and Intervention. 4th ed. Philadelphia, Pa: Lea & Febiger;1991:557-581.
Wynne J, Green L, Mann T, Levin D, Grossman W. Estimation of left ventricular volumes in man from biplane cineangiograms filmed in oblique projections. Am J Cardiol. 1978;41:726-732.
Lamas GA, Vaughan DE, Parisi AF, Pfeffer MA. Effects of left ventricular shape and captopril therapy on exercise capacity after anterior wall acute myocardial infarction. Am J Cardiol. 1989;63:1167-1172.
Sheehan FH, Bolson EL, Dodge HT, Mathey DG, Schofer J, Woo HW. Advantages and applications of the centerline method for characterizing regional ventricular function. Circulation. 1986;74:293-305.
Mitchell GF, Lamas GA, Vaughan DE, Pfeffer MA. Left ventricular remodeling in the year after first anterior myocardial infarction: a quantitative analysis of contractile segment lengths and ventricular shape. J Am Coll Cardiol. 1992;19:1136-1144.
Chesebro JH, Knatterud G, Roberts R, Borer J, Cohen LS, Dalen J, Dodge HT, Francis CK, Hillis D, Ludbrook P, Markis JE, Mueller H, Passamani ER, Powers ER, Rao AK, Robertson T, Ross A, Ryan TJ, Sobel BE, Willerson J, Williams DO, Zaret L, Braunwald E. Thrombolysis in Myocardial Infarction (TIMI) trial, phase I: a comparison between intravenous tissue plasminogen activator and intravenous streptokinase. Circulation. 1987;76:142-154.
Cox DR. Regression models and life-tables. J R Stat Soc (B). 1972;34:187-220.
Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457-481.
Heikkila J. Mitral incompetence complicating acute myocardial infarction. Br Heart J. 1967;29:162-169.
Bhatnagar SK, Yusuf ARA. Significance of a mitral regurgitation systolic murmur complicating a first acute myocardial infarction in the coronary care unit: assessment by colour Doppler flow imaging. Eur Heart J. 1991;12:1311-1315.
Perloff JK, Roberts WC. The mitral apparatus. Circulation. 1972;46:227-239.
Vokonas PS, Gorlin R, Cohn PF, Herman MV, Sonnenblick EH. Dynamic geometry of the left ventricle in mitral regurgitation. Circulation. 1973;48:786-795.
Gould KL, Lipscomb K, Hamilton GW, Kennedy JW. Relation of left ventricular shape function and wall stress in man. Am J Cardiol. 1974;34:627-634.
Kono T, Sabbah HN, Stein PD, Brymer JF, Khaja F. Left ventricular shape as a determinant of functional mitral regurgitation in patients with severe heart failure secondary to either coronary artery disease or idiopathic dilated cardiomyopathy. Am J Cardiol. 1991;68:355-359.
Sabbah HN, Rosman H, Kono T, Alam M, Khaja F, Goldstein S. On the mechanism of functional mitral regurgitation. Am J Cardiol. 1993;72:1074-1076.
Estes EH, Dalton FM, Entman ML, Dixon HB, Hackel DB. The anatomy and blood supply of the papillary muscles of the left ventricle. Am Heart J. 1966;71:356-362.
Reinfeld HB, Samet P, Hildner FJ. Resolution of congestive failure, mitral regurgitation, and angina after percutaneous transluminal coronary angioplasty of triple vessel disease. Cathet Cardiovasc Diagn. 1985;11:273-277.
Shawl FA, Forman MB, Punja S, Goldbaum TS. Emergent coronary angioplasty in the treatment of acute ischemic mitral regurgitation: long-term results in five cases. J Am Coll Cardiol. 1989;14:986-991.
Leor J, Feinberg MS, Vered Z, Hod H, Kaplinsky E, Goldbourt U, Truman S, Motro M. Effect of thrombolytic therapy on the evaluation of significant mitral regurgitation in patients with a first inferior myocardial infarction. J Am Coll Cardiol. 1993;21:1661-1666.
Lehmann KG, Francis CK, Sheeha FH, Dodge HT. Effect of thrombolysis on acute mitral regurgitation during evolving myocardial infarction. J Am Coll Cardiol. 1993;22:714-719.
Maisel AS, Gilpin EA, Klein L, Le Winter M, Henning H, Collins D. The murmur of papillary muscle dysfunction in acute myocardial infarction: clinical features and prognostic implications. Am Heart J. 1986;112:705-711.
Croft CH, Lipscomb K, Mathis K, Firth BG, Nicod P, Tilton G, Winniford MD, Hillis LD. Limitations of qualitative angiographic grading in aortic or mitral regurgitation. Am J Cardiol. 1984;53:1593-1598.